Summary The de HavillandDHC8 (registrationC-FACF, serial number259) was being operated by Jazz Air Inc. as FlightJZA7956 from Toronto, Ontario, to Windsor Locks, Connecticut (KBDL), U.S.A. The aircraft departed from Runway06L at Toronto/LesterB. Pearson International Airport (LBPIA), Ontario, at 1950 eastern standard time on the scheduled night instrument flight rules flight. During the take-off run, a three-foot piece of the wing leading edge (with the de-ice boot attached) separated from the left wing. The flight crew noticed a vibration on the flight controls during the initial climb after take-off and elected to return to Toronto/LBPIA. Air traffic control received a report of debris on Runway06L, and the DHC8aircraft leading-edge piece was retrieved. The aircraft landed without further incident on Runway 06L, with emergency response vehicles on standby. When the leading-edge section was examined, it was determined that the 14screws that secure the leading-edge section to the bottom of the wing were missing. Ce rapport est galement disponible en franais. Other Factual Information History of the Flight The flight crew arrived in Toronto, Ontario, from Cleveland, Ohio, at 1826 eastern standard time1 in a DHC8-100 aircraft and planned to depart at 1925 for Windsor Locks, Connecticut, U.S.A., as Flight7956. However, an earlier flight to Windsor Locks had been cancelled, and the passengers from that flight had been transferred to Flight7956. The increased passenger load necessitated a change to a larger aircraft, a DHC8-311. One consequence of changing aircraft for the return trip to the U.S.A. was that the flight crew now had to pass through U.S. Customs and Immigration. This was not problematic, but it was time consuming, and there was no extra time if the crew was to meet the scheduled departure time. While the captain checked the necessary paper work for the flight, including the weather, flight plan and aircraft maintenance records, the first officer completed a walk-around inspection of the aircraft. Photo1. Missing number 4, left, leading-edge panel The aircraft was parked on the East Satellite Terminal ramp. Flood lights illuminated most of the ramp area, but they also created dark shadows under the aircraft wings. The first officer required a flashlight to inspect the aircraft; the inspection of the aircraft, including an examination of both wings, revealed no anomalies. Maintenance History Two days before the occurrence, on 22October2002 at 0600, the day-shift maintenance crew began their shift. The position of hanger crew chief is not a permanent position, but is decided daily by seniority. A senior aircraft maintenance engineer (AME) accepted the position of crew chief for that day. Although the responsibilities of the crew chief are not well documented and no formal training is provided, the general understanding is that the crew chief is the liaison between the maintenance supervisor and the crew and is responsible for assigning and supervising the crew's work. In this instance, the crew chief's maintenance expertise was avionics, and there was a regional jet aircraft in the hanger with an avionics snag. The crew chief was working on this snag while carrying out his crew chief duties. At 0600, C-FACF was outside the hanger with a nosewheel steering snag. This snag was expected to be rectified quickly, so C-FACF was parked at the front of the hanger where it could be taken out with minimum disruption to hangar operations. After trouble shooting the snag, it was decided that both nose wheels would be replaced. While the crew chief was trouble shooting the avionics snag on the regional jet, maintenance control assigned an additional work package to C-FACF that included both a line check and an Out of Phase-3 (OPT-3) check. This work package was brought to the hanger floor by the maintenance supervisor and handed directly to the maintenance crew. One of the junior AMEs (AME1), who had been a licenced AME for three months and had been hired by Jazz Air Inc. less than one month prior to the occurrence, was tasked with showing some of the most recently hired AMEs how to complete the line check. AME1 had minimal experience on DHC8 maintenance and did not hold a company Aircraft Certification Authority (ACA). It was left to the crew to determine among themselves who would be responsible for each task on the OPT-3 check. The crew chief was not involved in the work assignments, but he was subsequently informed by the maintenance supervisor of the work package that had been handed out. The OPT-3 check consists of 11individual tasks. One of these was task3010/08, Operational Check of Pneumatic Equipment Heating System, which is a functional check of the electric heaters in the de-ice system. AME1 volunteered for this task, although he had no previous related experience. There are 11 steps on the de Havilland Inc. Dash8 Maintenance Task Card to complete task3010/08, which include removing the access panels, checking the six cockpit-controlled heaters, checking the seven thermostatically controlled heaters and reinstalling the access panels. There is provision on the task card for two signatures: one from the mechanic who did the job and one from the inspector who confirmed the job was completed. However, on the Transport Canada approved Air Canada Regional Task Card, task3010/08 has been simplified to a single line on the OPT-3 check. It requires only the initials of the individual who did the inspection. The entire OPT-3 task card, when complete, requires a single signature by an AME who holds a company ACA. Jazz Air Inc. uses the Air Canada Regional Task Card. AME1 was briefed by the crew chief and reviewed the maintenance instructions before he started task3010/08. He set up a work stand to get up to wing level, approximately 12feet above the floor, then removed the screws from the number4 leading-edge panel of the right wing. When the screws were removed, the leading edge was still secured by Product Research and Chemical Corporation (PRC) sealant. Without completing work on the right wing, because he was unsure of how to complete the removal of the leading edge, AME1 moved the work stand and tools to the left wing and started removing the screws for the left, number4, leading-edge panel. The right wing was not marked or flagged to indicate that the screws had been removed and that the work was incomplete. At approximately 1700, AME1 received voluntary help from an apprentice AME who was working on the afternoon shift and who had previously worked on a task3010/08. He informed AME1 that the PRC sealant had to be cut to release the leading edge and, if the sealant was cut cleanly, it could be resealed later. AME1 cut the sealant on the top surface of the left wing leading edge, but not cleanly; the apprentice AME then demonstrated how to do it on the bottom surface. They were then joined by another AME (AME2), who had been hired a week before. AME2 had not been assigned to a specific task, but he volunteered to assist. AME2 was experienced on other turbo prop and turbo jet aircraft but not on the DHC8 aircraft and he did not have the company ACA. Together, they removed the left leading edge and performed the heater checks on the left wing. AME2 then left to get some replacement PRC sealant and returned as AME1 finished re-installing the screws that secure the top of the leading-edge panel to the wing. AME2 and the apprentice AME then began applying the sealant to the top surface of the left wing leading-edge panel, and AME1 left to check the heaters in the tail of the aircraft before the end of his shift. The apprentice AME and AME2 were still applying sealant to the top surface of the leading edge when AME1 returned. He gave the apprentice AME the screws for the right wing leading edge and informed him the tail heaters were functional. Since the right wing heaters had not been tested, AME1 did not sign that task3010/08 was complete. AME1 then left for the day. The screws had not been installed on the left wing leading-edge panel, and it had not been sealed. The apprentice AME and AME2 moved the work stand to the right wing to complete that part of the job. As they finished cutting the sealant on the right wing leading-edge panel, the maintenance supervisor called the apprentice AME away from the aircraft to send him to the ramp. The apprentice AME repositioned the right wing leading edge and installed one screw so the leading edge would not fall. He then put the remaining screws in a latex glove, taped it to the leading-edge section and left for the ramp. AME2 also left, as his day shift was over. When the crew chief reviewed the work sheets before going home at 1800, he noted that task3010/08 had not been signed out. He had previously talked with AME1 and believed that he had checked all 13heaters. The crew chief initialled task3010/08 as being complete, even though some tasks had not been completed, because he did not want the night crew to have to recheck all 13heaters. The night shift crew arrived at 1930. After reviewing the aircraft paper work, they determined that, although task3010/08 had been signed off, the work was not complete. As well, the aircraft still needed an engine run to functionally check the de-ice boot operation, and the aircraft needed a taxi test to complete the rectification of the nosewheel steering snag. One of the night shift AMEs (AME3) met with the apprentice AME for a status report on task3010/08 when he returned to the hanger for his lunch break at 2000. His report was that the left wing was complete but the right wing was not. Two AMEs from the night shift crew examined the number4 leading-edge panel on the right wing and installed the remaining screws. The aircraft was then functionally checked, signed out and parked outside of the hanger as a hot spare. At this time, there were no screws in the bottom of the left wing leading-edge segment, it had not been sealed and the heating elements in the right wing leading edge had not been checked. Photo2. Cross-section of the left leading edge The aircraft sat all day Wednesday and Thursday, until Thursday night when it was assigned to Flight7956. No work card had been generated to seal either the bottom of the left wing leading-edge section or the right wing leading-edge section. The missing screws from the bottom of the left wing leading edge were not located. The number4 leading-edge panel is approximately 38inches long, Dshaped, with 11inches between the top and bottom rows of screws. The leading-edge panel is painted black and has a black pneumatic de-ice boot across the leading edge. It is secured to the wing structure by 14brass-coloured, countersunk screws on the top, and another 14on the bottom. Although there is some paint in the countersunk recess of the leading edge, most of the screw holes have a significant amount of bare aluminium showing. All of the screws securing all of the leading-edge sections are unpainted. Photo3. Left leading-edge section Other Occurrences On 04 January1998, a deHavilland DHC8-100, registrationN881CC, operated by CCAir, Inc., lost the number1 right-hand leading edge on departure from Charlotte, North Carolina. The leading-edge panel had been removed for maintenance and the bottom screws had not been re-installed. The flight crew was able to land the aircraft without further incident.2 On 11September1991, an Embraer120, operated by Continental Express, lost the leading edge of the left horizontal stabilizer over Eagle Lake, Texas. The loss of the stabilizer leading edge resulted in an in-flight breakup of the aircraft and the loss of 14lives. Forty-seven screws had not been installed, as required, in the top surface of the leading-edge panel.3